Use of New High Spatial Resolution 3T MRI Technique to Identify Prostate Cancer Missed on Routine Biopsy
David F. Yao, MD, William C. DeWolf, MD, Martin G. Sanda, MD, B Nicholas Bloch, MD, Elizabeth M. Genega, MD, Alexander M. Berry, MD, Long Ngo, MD, Neil M. Rofsky, MD, Neil M. Rofsky, MD.
Beth Israel Deaconess Medical Center, Boston, MA, USA.
Urologists are frequently presented with the dilemma of a patient who has had two or more negative prostate biopsies yet continues to have an elevated PSA and/or abnormal digital rectal exam. Cancer detection on repeat biopsy after two negative biopsies by sextant biopsy and contrast-enhanced ultrasound targeted biopsy are 5-14% and 16-20% respectively. We evaluated whether a novel MRI protocol can improve cancer yield, determined if detected tumors were clinically relevant, and compared MRI prompted biopsy results to those obtained with a concurrent systematic biopsy strategy.
A prospective cohort of 1053 patients undergoing prostate MRI at our institution was retrospectively interrogated to identify men referred for MRI after at least two negative biopsies. All men had PSA ≥4 ng/mL, PSA velocity ≥0.75/yr, or prior equivocal histopathology. Applying a novel T2-weighted and high-spatial-resolution dynamic contrast-enhanced protocol, imaging was performed with a 3-Tesla MRI and endorectal coil. Prediction and localization of cancer by MRI were assessed by an extended post-MRI biopsy that included a systematic survey (range 6-20 cores) and MRI directed biopsies (range 2-14 cores), simultaneously performed under TRUS guidance.
We identified 38 study patients with at least two negative biopsies (range 2-8) who underwent MRI and subsequent transrectal ultrasound guided biopsy. Mean age was 65 years (48-79); mean PSA was 14 ng/mL (2-33); and mean number of cores on the immediate pre-MRI biopsy was 16 (6-24). Twelve tumors were located by MRI prompted biopsies while two were detected by systematic biopsy. Of the tumors found by MRI, 75% were in the anterior lobe or central zone of the prostate and 93% were clinically significant (≥Gleason 7 or >50% of cores positive for cancer). The positive predictive value and positive yield of MRI were 55% and 32%, respectively.
MRI prompted biopsies resulted in a 6-fold increase for tumor yield compared to concurrent systematic biopsy. The anterior lobe and central zone, areas characteristically under sampled by systematic biopsies, contained a majority of these tumors. By Epstein’s histopathology criteria, the preponderance of identified tumors were clinically significant. Our results suggest that MRI prompted biopsies benefit patients with two or more prior negative biopsies by increasing positive yield of clinically relevant cancers.